Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
dandanua
searching Neon…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
6 ms
·
1.
▲
by
dandanua
6d ago
The next model will probably run simulation of a Universe to get a command output
2.
▲
by
dandanua
8d ago
Can't wait for the moment when AGI realizes how stupid and dishonest its owners are.
3.
▲
by
dandanua
8d ago
Your coworkers, after they learned about major progress in this problem, asked a model which was trained on the year of private work (the blog post even acknowledges this). No wonder it found the proof in less than a week using significantl
4.
▲
by
dandanua
8d ago
They have messed things up, because Levent has a conflict of interest between his job at Anthropic and this independent work, and Tristan should have opted out of OpenAI training on their work (he probably didn't know about this). This
5.
▲
by
dandanua
8d ago
Can you call paranoia a fear of something which is happening? OpenAI uses user chats for training and they are "open" about it.
6.
▲
by
dandanua
8d ago
Attempts to steal $1,000,000 for a solution to Millennium Prize Problem have become a tradition, apparently.
7.
▲
by
dandanua
9d ago
The nest step would be to map every human. Shouldn't be too hard, there are only 8B of us.
8.
▲
by
dandanua
12d ago
I wouldn't call it corruption when the richest man decides how the government's money should be spent. He is not an official, after all. I would call it a dead democratic system. EOL.
9.
▲
by
dandanua
13d ago
The Julia type system is also great for AI, it makes code generation more robust. I also liked that some Julia code (such as matrix multiplication) can be executed on a GPU with no change and will be almost as fast as possible, although com
10.
▲
by
dandanua
29d ago
I find his submission very valuable. It is not only digestible for humans but also summarizes the proof, explains what humans were missing, and highlights the methods that could help with stronger and similar conjectures (see his comments t
11.
▲
by
dandanua
1mo ago
LLMs don't understand things as human mathematicians do, even though they are very good at finding analogies and similarities. Their advantage is a larger search space (experience) and search speed, not better understanding.
12.
▲
by
dandanua
1mo ago
ancestral values coming back
13.
▲
by
dandanua
1mo ago
more likely alien race
14.
▲
by
dandanua
1mo ago
I'm sure US billionaires will find a way to extract those trillions from the public. They're smart, they can handle it. After all, they can ask AI for advice on how to do it.
15.
▲
by
dandanua
2mo ago
In other words, the US gov will simply put all the debt onto the public, as usual.
16.
▲
by
dandanua
2mo ago
But we now have trillionaires, aren't you happy about that? It means you might be the one some day too /s
17.
▲
by
dandanua
3mo ago
Nebula by Slack is also a decent player in this company.
18.
▲
by
dandanua
3mo ago
So, we live in the economy where everything is done by labour that receive minimal acceptable wages, while all the profits are collected by the ruling class who do nothing except constantly kissing each other asses.
19.
▲
by
dandanua
3mo ago
Since Trump is in charge, there is a high chance that this global stock market casino can go bankrupt as well. He is having a hand of Midas that works in reverse.
20.
▲
by
dandanua
4mo ago
My purpose was to say that humans might not be capable enough to see what's there beyond just rocks and gases. Actually, this should be a popular view, because, according to modern theory, the observable universe accounts for only 5% o
21.
▲
by
dandanua
4mo ago
> the rest of the universe thus far is just uncaring rocks and gases This is probably what ants think about humans when they run through their feet.
22.
▲
by
dandanua
4mo ago
This blog post clearly targets VCs, but what they are doing is legit and can improve the performance of local models on low-end hardware as well, especially since their priority is to optimize non-batched inference.
23.
▲
by
dandanua
4mo ago
The fact that users are using more search queries means they can't find what they want with a lesser number of queries. It seems that Google's PR team doesn't have an incentive to understand that, or thinks that everyone else
24.
▲
by
dandanua
4mo ago
The intention is to kill the web in its current form, obviously. If only 1/3 of their users have left, then it is still a win for them in the long run, as they will gain the fraction of content they directly supply to users. Singularit
25.
▲
by
dandanua
4mo ago
Well, this is what's happening in the stock market with the insider trading, pump and dump schemes, shorting and many more algorithms that utilize power imbalance to extract more capital for decades. Polymarket is quite tame in compari
26.
▲
by
dandanua
4mo ago
Dude, try not eating for a week and then publish your humble opinion on how good any change is.
27.
▲
by
dandanua
4mo ago
We are approaching the pinnacle of "privatize the profits, socialize the costs". No wonder the US is ruled by the straight fascists today.
28.
▲
by
dandanua
4mo ago
The story of Flash Attention is the best manifestation of power and difficulty of GPU programming. This page gives a nice overview of it https://aiwiki.ai/wiki/flash_attention
29.
▲
by
dandanua
4mo ago
You can also write CUDA kernels directly in Julia using CUDA.jl. I basically learned CUDA programming by experimenting in Julia with the help of LLMs.
30.
▲
by
dandanua
4mo ago
The paper https://eprint.iacr.org/2025/1296
More ›